The present study assessed the effects of amount of practice and length of delay on the learning and retention of the temporal motor sequence task (TMST). Participants learned to reproduce ten-element visual sequences by tapping in synchrony with the stimulus. Participants were randomly assigned to a varied-practice condition (n = 28) or a varied-delay condition (n = 40). Participants in the varied-practice condition received either 1, 3, or 6 blocks of practice on the TMST, on each of five consecutive days, followed by a fixed 4-week delayed-recall. Participants in the varied-delay condition received 3 blocks of practice on the TMST, on each of five consecutive days, followed by a varied delayed-recall of either 3 days, or 2, 4, or 8 weeks. Learning was assessed by changes in accuracy, response variance, and percent response asynchrony. Results showed that amount of practice had no significant effects on learning and retention of the TMST, suggesting that minimal amounts of practice spread over several days are sufficient to induce long-term memory of a motor skill. Delay appeared to differentially affect retention of the TMST, as length of delay influenced response accuracy, delay affected response synchronization, and neither delay nor length of delay had effects on response variance. These results indicate that different aspects of a motor skill are stored in independent but parallel systems. We propose that level of proficiency, rather than amount of practice or length of delay, is the critical factor affecting motor skill learning and retention.